You are on page 1of 10

6/22/2022

A NETWORK IS -

A network is a set of nodes and collection


of potential links between these nodes

Actors (nodes, vertices)


○ People, groups, organizations,
communities, nation-states, web sites,
documents, tags …
SESSION – 3: TYPES OF SOCIAL
NETWORK
PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 1 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 2

Relations (links, ties, edges,arcs)


○ Evaluations of one person by another (friendship, liking, ...)
TYPES OF NETWORK &
○ Transfers of material resources (lending, donations, ...)
ADJACENCY MATRIX
○ Association or affiliation (membership, attendance, …)
○ Behavioral interaction (communication, intercourse, …)
○ Movement between places (migration, mobility, …)
○ Physical connections (roads, routers, …)
○ Formal relations (authority, supply chain, …)
○ Biological relations (kinship, descent, …)
○ Retrieving and publishing documents
○ Tagging Photos

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 3 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 4

1
6/22/2022

DIRECTED VS UNDIRECTED
Component- A component is the largest subset of a network where all nodes are
directly or indirectly connected.
Communication vs
Friendship Network

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 5 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 6

NETWORK VISUALIZATION WITH R


https://kateto.net/network-visualization

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 7 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 8

2
6/22/2022

PURPOSE OF THE STUDY

Purpose: What decisions do you need to


make? Who is your audience and what do they
need to know?
Questions: What are you trying to measure?
´ What is the scope of your network?
 ´ Socio-centric
 ´ Ego-centric

´ What kind of connections do you want to know


about?
 ´ Undirected
 ´ Directed

PROCESS OF CONDUCING SNA ´ What attributes do you want to explore?


 ´ Gender
 ´ Geographic location
 ´ Etc.
PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 9 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 10

1. DATA COLLECTION – FOR EGO-CENTRIC NETWORKS DATA COLLECTION METHOD: SOCIO CENTRIC NETWORK
Survey Methods
Ask respondents (ego) to provide a list of Interviews
contacts.
Behavioral data
Web data
 Frequently asked questions: Who are your
best friends? With whom you exchange Archival data
resources and information most often? MyPersonality Data: Facebook ego networks and “Big Five” personality traits
(survey/archival/online).

 These contacts:
friends, relatives, acquaintances, strangers,
peers, team members and so on.
The Network Data Repository is a good place to look at, or you can build a network based
on public data (e.g., Twitter).

a look at NodeXL for public Twitter data.


PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 11 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 12

3
6/22/2022

METRIC USEFUL IN SNA:


1. SIZE OF THE NETWORK
2. ANALYSIS OF DATA

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 13 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 14

2. DENSITY 3. CENTRALITY
Not all nodes in a network are equally important in determining the network’s structure.

Density: describes the general level of


centrality indices provide insight into the relative importance of a node in the context of the other
linkage among the points in a graph. nodes in the network (Borgatti, 2005; Freeman, 1978).

It is a measure of number of potential Highly central nodes play a key role of a network, serving as hubs for different network dynamics.
ties that actually exist and the measure However the definition and importance of centrality might differ from case to case, and may refer to
of the average percentage of the different centrality measures:
network that an individual perceives as
being accessible by them. Centrality measures address the question: "Who is the most important or central person in this network?“
Degree centrality
Closeness centrality
Density = Actual Connection / Potential  Betweeness centrality
Connection
Eigen centrality
PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 15 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 16

4
6/22/2022

Degree — the amount of


neighbors of the node

Betweenness – the amount of


short path going through the
node

Closeness – the level of the


closeness to all of the nodes

EigenVector / PageRank —
iterative circles of neighbors
PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 17 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 18

DEGREE /CONNECTOR
Who is most important in this network? Degree centrality - in terms of in-degree and out-degree of a particular
node in a network.

John: Connector, John has the highest


degree in the network • Degree: total number of links with other nodes/actors.
In-degree- incoming links to a node/actor from other nodes/actors.
Out-degree: outgoing links from a node/actor to other notes/actors.
▪ For node j
In-degree = Σi A(i,j)
Out-degree = Σi A(j,i)

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 19 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 20

5
6/22/2022

CLOSENESS CENTRALITY / PULSE TAKERS BROKER


The pattern of Sue and Sam’s direct and Even though Jess has fewer connections than
indirect relationships allow them to average, her position has great influence
access all other nodes faster than
anyone else
Jess is a ‘broker;’ she has high betweenness
centrality and controls the group’s information
They are ‘pulse-takers’ (have high flows
closeness centrality) and can monitor
information flow better than others

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 21 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 22

USEFUL METRIC IN ANY SNA:


Number of network members

Diversity of network members (e.g. different types of actors in the network)

Connectedness (e.g. percentage of network members connected or linked to each other)

Reach (the number of steps or path lengths between different organisations in the network)

Reciprocity (the proportion of interactions that are mutual, e.g. where information flows in both directions)

Number of clusters (e.g. clusters where members have much interaction with each other but little connection with other
network members)

Degree of centralization (the degree to which relationships are dominated by one organisation or a group of organizations)

Sustainability (which can be calculated by looking at the potential for a network to fragment if one or more members
leaves or becomes inactive)
PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 23
Outreach (how many alliances members have relationships and connections outside of the
PREPARED BY DR PREETI KHANNA FOR SNWA COURSE
network) 24

6
6/22/2022

3. VISUALIZE DATA
4. INTERPRETATION OF DATA
So what?
´ How does it relate to what you’re trying to measure?
´ What does it mean for you ?

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 25 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 26

To complete the social network analysis, software packages will be needed to complete the following
tasks: SOCIAL NETWORKING TOOLS: COMPARISON
Data collection – E.g. Spreadsheet software;
Data analysis – E.g. Social network analysis software;
Data visualization – E.g. Network visualization software.

Some of the most often cited free packages are UCINET/NetDraw, Gephi and NodeXL

Links to the three SNA software packages mentioned can be found below
• UCINET/NetDraw - https://sites.google.com/site/netdrawsoftware/home
• Gephi - https://gephi.org/
• NodeXL - https://nodexl.codeplex.com

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 27 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 28

7
6/22/2022

SOCIAL NETWORKING TOOLS: PROS & CONS MUST WATCH


 Gephi: visualization and basic network metrics
 iGraph: for Programming assignments
 NetLogo: modeling network dynamics
 NetworkX (Python): open source, extensive functionality
 Netlytic: cloud-based text and social networks analyzer
 NodeXL (Windows only): SNA integrated into Excel
 Pajek (Windows only)
 SNA package for R
 SNA in Python
 SoNIA: social network image animator specialized for longitudinal analysis of networks
 UCINet (Windows only): sociology-focused functionality
 SNAPP: Social networks adapting pedagogical practice
 Condor: The goal of this research project at the MIT Center for Collective Intelligence is to help
organizations to increase knowledge worker productivity and innovation, by creating Collaborative
Innovation Networks (COINs).
https://ieeexplore.ieee.org/document/9318162

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 29 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 30

STEP 1: REPRESENTING RELATIONS AS NETWORKS

EXAMPLE OF SNA
PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 31 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 32

8
6/22/2022

REPRESENTING AN UNDIRECTED GRAPH

ENTERING THE DATA INTO DIRECTED GRAPH

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 33 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 34

EGO NETWORKS AND ‘WHOLE’ NETWORKS


ADDING WEIGHTS TO EDGES
Weights could be –

Frequency of interaction in period of observation


•Number of items exchanged in period
•Individual perceptions of strength of relationship
•Costs in communication or exchange, e.g. distance
•Combinations of these

* no studied network is ‘whole’ in practice; it’s usually a partial picture of one’s real life networks (boundary specification problem)
** ego not needed for analysis as all alters are by definition connected to ego PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 35 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 36

9
6/22/2022

CASE 1: IMPROVED KNOWLEDGE-SHARING FOR A GLOBALLY OPERATING CONSUMER TECHNOLOGY CLUSTERING AND KNOWLEDGE SHARING WITHIN THE CONSUMER GOODS COMPANY
GOODS MANUFACTURER

For a major consumer goods manufacturer, innovation centers across the world, with
many Research & Development (R&D) employees working together at each location.
Due to the sheer size of the organisation, it was hard for employees to actually locate
each other, and thereby each other’s knowledge.
Essentially, knowledge was sourced mostly within the same divisions. At the same time,
our client realized that cross-disciplinary knowledge sharing could benefit the innovation
process significantly.
Using a mix of data sources in the context of innovation such as scientific publications,
patents and conference proceedings, information on collaborations which had taken
place within the organisation during a 5-year time-frame was gathered.

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 37 PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 38

MUST WATCH

PREPARED BY DR PREETI KHANNA FOR SNWA COURSE 39

10

You might also like